This application is in response to: 08-HL-101* "Identify causal genetic variants associated with heart, lung, and blood diseases by application of targeted DNA capture and massively parallel sequencing technologies followed by selective genotyping of DNA samples from large well-phenotyped populations." Plasma lipids including low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C) and triglycerides (TG) are highly heritable risk factors for cardiovascular disease. We and others have used genome-wide association (GWA) mapping in populations to define many new genomic loci related to plasma lipids. However, our knowledge of the genetic basis for plasma lipid concentrations remains incomplete. For mapped loci, we need to move from locus to gene and pinpoint the specific causal gene responsible for the association. We need to address allelic heterogeneity, the possibility that multiple alleles within a single gene contribute to trait variation. And perhaps most importantly, we need to evaluate if the newly discovered loci not only relate to plasma lipid concentrations but also confer risk for clinical CVD. Specifically, we hypothesize that: (1) each region identified by GWA studies (GWASs) contains at least one gene causal for plasma lipid traits;(2) in addition to the common variant tagged in the initial GWAS, the causal gene contains one or more additional common (>5% MAF), low frequency ((0.5% - 5% MAF) and/or rare mutations (<0.5% MAF) that influence plasma lipid traits;and (3) that variants at some loci will relate not only to plasma lipids but also risk for CVD. To test these hypotheses, we have assembled four resources: (1) the Global Lipids Genetics Consortium - a GWAS for LDL-C, HDL-C, and TG in >100,000 participants;(2) subjects with very extreme lipid phenotypes ascertained in referral lipid clinics;(3) a set of protocols and analytic methods for deep sequencing using next-generation technology;and (4) five population-based cohort studies with baseline lipid measures and longitudinal follow-up for incident CVD. We will identify additional common, low-frequency and/or rare mutations by sequencing 127 genes in the associated intervals at 26 loci defined by the Global Lipids Genetics Consortium as having association P <10-14 in subjects with extreme lipid phenotypes (>95th percentile for LDL-C, HDL-C, and TG;n=500 in each group) and controls for each group (<25th percentile;n=500 in each group). Sequencing in subjects with very extreme lipids should markedly enhance the ability to discover causal alleles and the probability of finding null alleles at causal genes. We will then perform a validation of newly discovered low frequency (0.5% - 5% MAF) non-synonymous variants in population-based cohort studies (n~35,000) and test whether these coding variants are associated with plasma lipid and lipoprotein traits and confer risk for incident CVD. Our research team has made important contributions to lipoprotein and CVD genetics, with a long track record of effective collaboration and leadership. The group is experienced and skilled in genetics, genomics, next-generation sequencing, statistical and population genetics, lipoprotein physiology and epidemiology. Using this skill set and the samples assembled, we hope to identify the specific causal allele(s) at 26 independent loci that we have mapped for plasma lipid concentrations and clarify which genes relate to risk for CVD, in the process providing important validation for specific genes as novel new therapeutic targets for CVD. We propose to perform DNA resequencing collected cohorts with extreme lipid phenoptypes (elevated LDL-C, triglycerides and HDL-C). We will resequence areas of the genome that have been identified as potential loci affecting lipid traits in genome wide association studies. By utilizing extreme lipid phenotypes we expect to identify new causal genetic variants, and will then test these variants for their relationship to cardiovascular disease.